Inter-observer reliability in the assessment of heterotopic ossification: proposal of a combined classification

We applied four published classifications for assessment of heterotopic ossification after total hip arthroplasty (Arcq, Brooker, DeLee and Della Valle [1, 2, 5, 6]. The average incidence of heterotopic ossification varied from 19.8% to 27.7%. The inter-observer reliability of the various classifications was determined by Cohen’s Kappa statistic. Kappa values ranged from 0.897 for Arcq’s to 0.814 for Brooker’s classification. In order to increase the reliability and consistency, we propose a new classification system combining Brooker’s and Della Valle’s classifications. This new classification preserves the high reliability of Della Valle’s system and is comparable to previous publications since it includes Brooker’s criteria.